Rees Burridge was born on 7 August 2006. Just short of 8 weeks later, on 30 September 2006, he was admitted to hospital suffering from irrecoverable injuries and, on 2 October 2006, life support was discontinued and he died. An investigation of the circumstances which led to his admission to hospital resulted in the prosecution of his father, Michael Burridge, for murder. On 4 March 2008, in the Crown Court at Plymouth before Royce J and a jury, after a trial lasting 12 days, he was convicted of that offence and sentenced to life imprisonment with a period of 13 years being specified as the minimum term to be served. Over 2½ years later, his application for leave to appeal against conviction (based upon grounds which, as we shall later explain, have twice been recast) falls to be decided; having regard to the history, we deal with all the grounds that have been advanced.

The Applicant was born on 10 August 1979 and, at the material time, was 27 years of age. Having served two apprenticeships, he joined the army at 17 and served until 2002 when he left with a certificate of service which spoke highly of him and described his record as exemplary. Also in 2002, he married Donna Burridge (whom we shall describe as Donna throughout) and their first child, a daughter, was born in July 2004. It was common ground that Donna was much affected by in the deaths of her parents just before and shortly after the birth of their daughter. The jury also learnt that, at or about the end of 2004, the Applicant’s behaviour (particularly to other drivers of whose behaviour he disapproved) led to him going on an anger management course although he only attended an assessment and one session. In 2005, he had an affair with Donna’s cousin; quite apart from the effect it had on Donna, it was said to be relevant because it had affected the minds of her relatives against him.

When Donna became pregnant with Rees, he told the jury that he was pleased because he thought it would improve their marriage which was clearly stressful, not only because of the affair, but also because Donna continued to suffer from depression and, additionally, suffered considerable back pain following a road traffic accident years previously. A reflection of the position might be seen from discussions that husband and wife had with a clinical psychologist. On one occasion, the Applicant apparently commented that it would be better for everyone if he was dead and that he had thought about using carbon monoxide in the garage (although he later said the remark was flippant); Donna also admitted to suicidal thoughts.

After that short introduction, we turn to Rees and an account of his short life, the significant events in which were conveniently identified for the jury in an agreed chronology which starts with his birth, at 9.42 on 7 August 2006, by elective Caesarean section at 37 weeks gestation. Although his condition was described as good, shortly after he was delivered he showed signs that he was struggling to breathe, and so oxygen was supplied. His condition soon improved but, for a time, he was transferred to the Neonatal Intensive Care Unit. A chest X ray showed no fracture of any rib. There was a small hole in his heart, which is not an uncommon finding and was not regarded by any witness at trial as significant. Over the following 24 hours, he did very well and was allowed to leave hospital on the 10 August. His birth weight was between the 91^st and 98^th centile and his head circumference at birth about the 98^th centile.

In the first 24 hours following discharge, he was seen at home by a health visitor and by a general practitioner, Dr Cullen. On 12 August Rees was readmitted to hospital because of jaundice and weight loss. He was seen by Dr Allwood, a consultant neonatologist, and discharged home on the basis that he would return the following day for review. Donna duly took him back to hospital the following day. Again, he was seen again by Dr Allwood, and again, he was discharged. It was suggested to Dr Allwood in cross examination that the jaundice and/or the breathing difficulties at birth might indicate the presence of subdural haemorrhage at the time of birth. For reasons which he explained to the jury, Dr Allwood rejected that suggestion.

On a number of occasions between 16 August and 1 September, Rees was seen at his home by at least three different health visitors. None of them noted anything abnormal about his condition. Both parents were observed handling their baby son in an appropriate manner. On or about 2 September, however, it is said that there was an incident involving Rees and his elder sister (who, it is important to note, was then just short of 26 months old). In evidence, the Applicant said that he was putting shopping away in the kitchen when he heard his wife shouting in the living room. When he went to see what was happening, Rees was on the floor and his daughter was looking scared. Donna told him that Rees had been held by his sister around the throat and, presumably when she shouted, “thrown” onto the carpeted floor so as twice to hit his head. He said that Donna picked Rees up; he was quiet “for a couple of minutes” and then given a feed. Both parents thought he was all right and so did not seek any medical attention. Donna (who was not a witness in the trial) subsequently gave a number of different accounts of this incident to different health professionals and members of her family, which were not consistent with one another: some were in the presence of the Applicant and a number were recounted in evidence, although none was accepted by the Crown as true. We shall return to the detailed account of the incident which Donna gave when interviewed by the police as a witness, on the basis that the failure to use it has formed a ground of appeal.

Returning to the evidence before the jury, the Applicant said that on the day following this incident, 3September, he and Donna noticed a little blood in one of Rees’s eyes. In addition, he said, Rees began frequent projectile vomiting and appeared to be in pain. Although noting these developments, medical attention was still not sought.

On 6 September, Donna saw a health visitor and reported that Rees was in pain and that popping noises could be heard in his chest. Later that day the Applicant, equipped with a note from Donna setting out her concerns, took Rees to the GP surgery. Rees was seen and examined by Dr Flanagan, who gave evidence about that examination. The Applicant reported to the doctor that Rees had been dropped on his head by his sister, and that since that event there had been popping or clicking noises emanating from his chest. On examination by Dr Flanagan, Rees appeared normal save for a small sub-conjunctival haemorrhage in one eye. Specifically, he showed no sign of discomfort when she was examining his ribs. Her evidence was that she was satisfied that Rees had not suffered any significant injury as a result of the incident with his sister. The Applicant himself, when interviewed by the police, said that after this visit to the doctor Rees was fine apart from the fact that he continued to vomit frequently.

On the following day, 7 September, Donna’s sister Janice Randall (herself the mother of young children) visited the house for supper. She gave evidence at trial describing Rees as being very quiet, and one of his eyes as being very bloodshot. She was given an account by Donna of the incident when Rees was dropped: Donna told her that Rees had been initially unresponsive and then very vacant for about 20 minutes. The Applicant, who was present during this conversation, appears not to have demurred from that account. Mrs Randall made a further visit to the house a week later, on 14 September: she noticed nothing unusual about Rees.

On the 19 September Donna took Rees back to the surgery, where he was seen by Dr Cullen (who had had contact with the family since 11 August). She told the doctor that Rees was unwell and that she was worried about his feeding because he vomited after food. Dr Cullen examined Rees, feeling his tummy and chest: no abnormality was disclosed and he appeared normal.

On 24 September, Rees was christened. As one would expect, a number of those present at the christening held him. He was described by witnesses as being pale, quiet and sleepy, but otherwise no witness reported noticing anything wrong with him. On 25 September, Rees spent some time in the care of the Applicant’s stepmother. She noticed that one pupil was larger than the other, and that he was pale and sleepy; she also reported that he showed no sign of discomfort or distress.

On the 26 September, Rees was seen again by Dr Cullen for the review which is routinely carried out at 6–8 weeks of age. Donna reported that frequent vomiting, sometimes projectile, was still continuing. Dr Cullen found no physical abnormality, but a hospital appointment was made for 28 September. That appointment was subsequently cancelled after Donna telephoned Dr Cullen and said that Rees was better and that the vomiting had settled.

We now come to 30 September, when plans had been made for Donna to spend the day Christmas shopping with her sister Janice. The Applicant was to look after the two children. With the children, he drove Donna to Janice’s house and after the two sisters had set off on their shopping trip, he remained for a time with Janice’s husband, Kevin. He then went off on an errand, leaving Rees in the care his brother in law. Kevin, who was aware that there had been reference to a problem with one of Rees’s eyes, told the jury that he made a point of looking into his eyes and moving his hand around to see if his eyes followed the movement. All seemed to be normal; the baby’s colour was okay and he seemed content. The Applicant, on his return, gave Rees a bottle and, again, according to Kevin, he seemed all right. The Applicant took Rees home at about 11 or 11.30am. Kevin Randall was the last person to see Rees before the events of the evening when he was taken to hospital. His evidence was therefore important as to Rees’s general condition only a matter of hours earlier.

In the course of the day, the Applicant sent a text message to his wife “All good here”. When she phoned at 3.40 pm, he said all was “OK”. At 7.44 pm he sent her another text message asking how far from home she was. Donna did not notice the arrival of that text until a few minutes later, and she then rang the Applicant. The call was timed in the call billing records at 7.55pm. In that call the Applicant told her that he was just about to drive to the hospital because Rees was not breathing, and appeared to be lifeless. She and her sister immediately diverted from their route in order to go straight to the hospital. They arrived at about the same time as the Applicant, and it was in fact Donna who carried Rees into the hospital at about 8.12pm.

The Applicant gave an account to the effect that Rees had passed the day uneventfully. Rees was supposed to have been given a feed at 5.00 pm but this slipped the Applicant’s mind. He left him in his cot during the day until about 7.00 pm when he undressed him for a bath. Rees had given a double intake of breath when first placed into the bath water (which he attributed to the water being a bit too cold), and had then been unusually quiet when being bathed, which he did not think anything of. He had looked all right and had been breathing normally when the Applicant was dressing him after the bath, which he did in day clothes which had just been bought for him. He put Rees into his cot and said that, after sending his 7.44 pm text message to his wife, he had gone to pick Rees up and found him pale with bluish lips. He then put him on the dressing unit and tried to find a pulse but could not. He tried cardio-pulmonary resuscitation (“CPR”) by breathing into him while holding his nose; he said he rubbed his tummy and tried to do compressions for about a minute. He still could not find a pulse.

It was then that the Applicant put Rees (and his daughter) into the car to drive to hospital rather than wait for an ambulance. He was just about to leave when Donna rang him in response to his text message. Thus, as a matter of timing, it was the Applicant’s chronology, maintained to the jury in his evidence, that the first time he had had any reason to be concerned about Rees was after he had sent his text message at 7.44pm, and therefore no more than about 30 minutes before he arrived at the hospital. On this account, nothing had happened which could explain why Rees had suffered such a sudden collapse.

At the hospital a staff nurse took Rees from his mother. He was blue, floppy and lifeless. The staff nurse started CPR, which she continued for about 15 minutes. This consisted of her placing her two fingers on his chest and compressing it on the nipple line. She estimated that she started that at about 8.14 pm, a couple of minutes after the first time recorded in the notes of 8.12 pm. She was asked whether her two finger compression could have been responsible for a small area of bruising in the centre of the chest (described by Professor Risdon at the post mortem as a grey apparent bruise (0.6 cm x 1 cm) present in the mid-line of the chest over the sternum and in line with the two nipples with was fresh haemorrhage in the subcutaneous fat). She said it would: indeed that was the small area of bruising found by Professor Risdon at post mortem. She did not know whether her CPR could have been responsible for the fracture of the ribs: she did not know, but she said she did not feel any sensation of ribs breaking during the procedure.

Later that night the Applicant was arrested at his home on suspicion of attempted murder. At that stage, Rees was still alive, his heart having been restarted by the resuscitation which had been carried out at hospital. When cautioned, the Applicant (referring to the incident witnessed by Donna) replied “Is it something my daughter could have done?” In his subsequent police interview he recounted the history of Rees’s short life, and the events of the 30 September. He denied that he had ever lost his temper, or had a violent outburst, with either of his children. He had never done anything to Rees which could have caused any broken rib. He said nothing untoward had happened on the 30 September before he found Rees was not breathing. He said that in trying to resuscitate Rees, he had rubbed vigorously on his chest and tapped his back in a way that was not vigorous (“I know he’s only little”); he pressed Rees’s rib cage with finger pressure (“I don’t think it went in very far”). He did not hear anything break and he did not think he had used enough force to break a rib. The evidence which the Applicant gave to the jury was consistent with what he had told the police in interview: he said that there was nothing in his efforts that would have damaged Rees’s ribs.

In the meantime, Rees had been transferred to a different hospital. There was no possibility of successful intervention, and on the 2 October life support was discontinued and death supervened.

In brief summary, that was the evidence before the jury as to the background. Before embarking on a consideration of the medical evidence, however, it is necessary to underline what was not before the jury. Thus, Donna was not a witness in the case. The Crown, as we have said, did not accept as true any of her accounts of the incident said to have involved her daughter, and did not regard her as a credible witness. She would of course have been a competent and indeed compellable witness for the defence, but was not called, although the jury did hear evidence, not only from the Applicant but from others, as to what Donna had said about that incident and about Rees’s subsequent state of health. Without objection, that evidence included pure hearsay in the form of witnesses reporting what Donna had said to them at times when the Applicant was not present. We can well understand why the Crown took the view that that evidence should all, in fairness to the Applicant, be before the jury. It did not, however, include reference to Donna’s interview by the police or to the DVD recording of that interview as a witness, which included a demonstration of what had transpired between Rees and his sister. The DVD was part of the unused material, and was provided to the defence: no application was made to play it to the jury. We return to this topic later in this judgment.

It is unnecessary to rehearse the medical evidence that was placed before the jury in relation to the efforts made to resuscitate Rees and care for him prior to his death, or the accounts provided to them by the Applicant and Donna. Suffice to say that the Applicant’s arrest followed as a result of the conclusion reached, from evidence of head injury to the brain and eye (the ‘triad’ of signs) and rib fractures, that Rees had suffered two episodes of trauma which, in the absence of an adequate history to account for them, were non-accidental. Those who examined Rees after his death provided considerable elaboration, and in the light of the present challenges to the Applicant’s conviction, we deal separately and comparatively extensively (but not exhaustively) with the evidence in both of these areas.

On 30 September, soon after his admission to Derriford Hospital, Rees was subject to CT scans of his skull which were reviewed by Dr Neil Stoodley, a Consultant Neuro-Radiologist with a special interest in paediatric brain injuries. He noted a number of subdural bleeds showing up as bright material on the scans and indicating that the bleeds were recent. There was recent blood in the inter-hemispheric fissure on both sides at the back of the head, on the right side of the inter-hemispheric fissure at the front, in the posterior fossa at the back of the head at the bottom, and possibly a small amount of recent haemorrhage over the right frontal lobe. Dr Stoodley concluded that the pattern of recent subdural haemorrhage on these scans was most likely to have been caused by head trauma. There was no evidence of any naturally occurring medical condition which could explain the pattern of bleeding.

There was also evidence of bright recent blood at several separate sites within the subarachnoid space, as well as in very small volume over the surface of the brain, in the convolutions of the brain where cerebro-spinal fluid normally sits within the lateral ventricles. The peripheral pattern of small amounts of subarachnoid haemorrhage was itself very typical of head trauma. The appearance of the brain itself showed extensive abnormality, the cerebral hemispheres being darker than they should have been on the scan, and the normal differentiation between grey and white matter largely lost. These appearances were indicative of very extensive hypoxic ischaemic brain injury. Further, the brain itself was not greatly swollen and the spaces within the head normally containing cerebro-spinal fluid were still present on the scan. Bearing in mind that the scans had been performed within about 3 hours of Rees’s collapse, Dr Stoodley concluded that the causative event was likely to have occurred very close to the time of Rees’s clinical collapse.

Further, the combination of the recent subdural haemorrhage at different separate sites, together with hypoxic ischaemic brain injury, was (according to Dr Stoodley) most likely due to an episode of head trauma. As to timing, the acute blood could not date back to the time of Rees’s delivery nor, given its appearance, was it likely to be related to his birth and he had, in any event, been delivered by caesarean section and so had not undergone the pressures of a vaginal delivery. The appearances meant that it was possible, although very unlikely, that the relevant event was severe accidental head trauma, but, again, there was no such history in Rees’s case.

Dr Stoodley also found evidence of darker subdural fluid over the left cerebral hemisphere. He identified two main possible explanations for that. On the one hand, it was possible that this represented evidence of an older episode of subdural bleeding, which had led to a chronic subdural haematoma. Dr Stoodley’s preferred explanation, however, was that damage to the arachnoid membrane had permitted cerebro-spinal fluid, which is dark, to leak into the subdural space and either collect there or dilute any acute blood present. Dr Stoodley understood that in this case the pathologists had found evidence of subdural membranes, and hence evidence of previous episodes of subdural bleeding, where the blood had resolved itself into healing membrane. Although he accepted that it was possible, he did not accept that this necessarily meant that the darker fluid over the left cerebral hemisphere was a chronic subdural haematoma.

For the purposes of his cross examination, Mr Paul Dunkels QC (who then appeared for the Applicant with Mr Robert Linford) was assisted by Dr Waney Squier, a Consultant Neuro-pathologist (who, for reasons to which we shall later refer, did not herself give evidence). Thus it was suggested to Dr Stoodley that the causation of the episode which killed Rees might have been a bleed from the chronic or old subdural haematomas. Dr Stoodley rejected as very unlikely this thesis of re-bleeding, making it clear:

A further proposition put to Dr Stoodley was that the fresh bleeding in the brain could have been a consequence of the leaking of blood from damaged blood vessels, when Rees’s circulation was restored, subsequent to hypoxic ischaemic damage in the 35 minutes between the Applicant noticing that Rees had stopped breathing and the recommencement of circulation in hospital. This came to be termed the “re-perfusion” theory. Intrinsic to this suggestion is the proposition that the hypoxic ischaemic encephalopathy arose from another cause, perhaps unknown. This aetiology for brain injury, said to explain the “triad” of encephalopathy, retinal bleeds and subdural haemorrhages, has come to be known as the “unified theory” or alternatively “Geddes 3”, after its original formulator Dr Geddes.

It is clear from the evidence before us that Dr Geddes has herself withdrawn her support for this theory, although (as is evident from her report prepared for the defence in this case and the record of a recent symposium at the Royal College of Pathologists, to which we will later refer) Dr Squier remains a proponent of this thesis as a potential explanation of the “triad”. It suffices to say that Dr Stoodley rejected this explanation as “extraordinarily improbable”. He told the jury that the:

We turn to the post mortem findings and to Professor Anthony Risdon who spent much of his career as a senior Paediatric Pathologist at Great Ormond Street, with a particular interest in sudden death in infants in their first year of life. Since his retirement from the NHS in 2004, he has practised as a forensic pathologist and, in that capacity, on 5 October 2006, conducted a post mortem examination on the body of Rees Burridge.

Professor Risdon’s findings in this area can be summarised as follows. He found a thin layer of subdural haemorrhage over the top of both brain hemispheres, round at the back and below the brain. There was a little sub-arachnoid haemorrhage. There was fresh fluid blood in the subdural space round the whole of the cord and right the way down to the bottom of the cord. Professor Risdon found haemorrhage – fresh red blood – around both optic nerves. The brain substance he found to be very abnormally soft. The fontanelle (that is to say the soft spot on top of the head where the skull bones of a baby have not yet fused), was quite tense to the touch, and the lines of fibrous tissue between the individual skull bones or plates, were considerably stretched, with a little haemorrhage within them. He told the jury this was an indication that the brain underneath was swollen and under tension. Professor Risdon believed that the triad of injury pointed to acceleration/deceleration of the brain and the skull which was non accidental: it was not just the fact of the three injuries but their nature and extent. As to timing, he said that if Rees looked normal and fed on 30 September, he did not consider that the act could have been before that.

Evidence in relation to the ophthalmic injuries came from Professor Luthert, the Professor of Pathology at the Institute of Ophthalmology at University College. An examination of Rees’s eyes revealed bleeding in both eyes around the optic nerve, which extended into surrounding tissue. Blood was seen between the optic nerve and the surrounding sheath of dura mater and, inside the eye, he found very extensive bleeding from the front of the retina to the back of the retina, which he described as “confluent”. On microscopic examination, he found cell evidence which suggested a time interval of two or three days between the bleeding and the point of death. Professor Luthert explained the significance of his findings in terms that they were most commonly seen in the context of alleged head injury. Given the other findings in this case, it was his opinion that the explanation for the retinal haemorrhages was much more likely to be trauma, than any of the alternative explanations.

To Mr Dunkels, Professor Luthert agreed that, beyond the fact that by definition in relation to cases in which he examined a body, the force was enough to cause death, it was unsafe to link the extent of retinal haemorrhage to the degree of force employed. He was pressed as to whether the retinal haemorrhages might not arise from brain damage caused other than by trauma, on the basis that the suggested sequence was as follows: (1) brain damage arising other than from trauma, (2) leading to brain swelling with raised intra-cranial pressure, (3) causing hypoxic damage to the very delicate blood-vessels in the eye; (4) such damage leading to a failure of auto-regulation when blood flow is restored, for example during resuscitation, meaning (5) that the vessels cannot adjust to the restored flow; and hence (6) bleeding occurs at the point of restoration of blood supply. This again is the “re-perfusion” theory. Professor Luthert accepted that such a mechanism might be an important component in some cases of retinal bleeding, and was a possible cause of retinal bleeding in this case, but in his opinion such bleeds did not occur “anywhere near as frequently as they do when the context is of alleged trauma”.

In that regard it is worth adding that Mr Markham, a Consultant Ophthalmic surgeon consulted after Rees’s admission, expressed a similar view. He said that loss of oxygen would not have been sufficient to cause these changes and only sudden and severe intracranial increase of pressure to the brain (such as throttling or some direct trauma such as a blow) would be sufficient: he thought shaking was very unlikely to have been the form of trauma.

Dr Brian Harding has been a Consultant Neuro-pathologist at Great Ormond Street since 1983, and is the only full-time paediatric neuro-pathologist in Britain. He also examined Rees’ brain and noted extensive haemorrhage in the subarachnoid space above the brain and below it. The brain was soft and the lowest part of the brain was necrotic or, as he told the jury, “in other words dead and breaking up”. Under the microscope, the brain tissue was “clearly very abnormal” with many little haemorrhages.

He also examined the spinal cord from within the back, where he found a lot of fresh haemorrhage around the outside and underneath the dural membrane. In addition, there was chronic subdural haemorrhage, meaning there had been an old bleed which was undergoing repair at a cellular level. There was subarachnoid haemorrhage in the spinal cord, and extending out of the cord itself into the surrounding tissues. This expert examined the spinal cord at chest level, approximating to the region of the ribs. He here noted acute and chronic (fresh and older) subdural haemorrhage and some hypoxic change in some nerve cells. Using a special stain, Dr Harding found some abnormally reactive axons/large nerve cells just where the nerves leave the cord at the “root entry zones”. He also observed a “jagged cavity running out from the central canal of the cord” at the level of the lumbar spine. He agreed this represented possible tearing of the tissue in the cord. Examination of the cranial area revealed a chronic subdural membrane, in other words a reactive process to an earlier haemorrhage, confirming that there had been an older bleed.

Dr Harding summarised his findings by saying he had found brain swelling, and further brain damage following the swelling, bleeding around the cord with suggested axonal damage and possible tearing to the tissue in the spinal cord. He noted old and recent rib fractures (to which we shall return), extensive retinal haemorrhages, encephalopathy and subdural haemorrhages and gave it as his view that the “triad of abnormalities” found suggested an acceleration/deceleration injury of some sort to the brain. He added that in this instance there was also evidence of a previous bleed in the brain and previous rib fractures, and concluded that his impression was there had been two times when brain damage had occurred in conjunction with other injuries. Without any clear evidence of something such as a bleeding disorder or severe accidental injury, the evidence suggests “some type of shaking incident”. The acute damage appeared to him very recent and he thought that:

In cross examination, Dr Harding accepted that bleeding at the point of re-perfusion of the brain was probably the cause of at least some of the damage within the brain. He said that it was possible that the old subdural haemorrhage around the spinal cord could have drained from beneath the dura around the brain although this did not alter the timing of the injury. His view was that the ‘jagged cavity’, running from the central canal in the lumbar cord with haemorrhage present within it, was not caused by hypoxic ischaemic damage but “was of a different nature than the other damage to the brain that I saw”. The brain tissue which was damaged had quite a different appearance and was much more severely damaged by lack of oxygen, “whereas this tissue was much more intact apart from this very curious discontinuity which I hadn’t seen before”. He could not put forward a definite explanation for this feature.

Mr Dunkels also cross-examined Dr Harding on his findings as to the damage to the nerve cells at the level of the cervical cord. The expert rejected hypoxic ischaemic damage as the cause of these injuries since:

It was suggested that these findings might derive from a shift of the cord down the spinal column as a result of brain swelling, but Dr Harding rejected that explanation. In the context of brain swelling, the shift down would be slow, whereas this damage required a shift which was much more rapid. He could not see how hypoxia would produce that damage. When pressed as to whether subdural bleeding might arise in the context of re-perfusion, he stated that subarachnoid haemorrhage is often seen in cases of hypoxia and re-perfusion but subdural haemorrhage is not normally seen in the context of hypoxia, although he acknowledged that there was “some divergence” of medical opinion on that point. Despite examining brains since 1983, Dr Harding made it clear that he had never seen this type of subdural haemorrhage associated with re-perfusion following severe hypoxic damage although he also found acute haemorrhagic infarction, the consequence of hypoxic damage, in most areas of the brain, and in particular at the back of the brain.

On 7 August 2006, shortly after Rees was born, there was an X ray which revealed no injury to any of his ribs. A further X ray on 30 September (and skeletal surveys on 2 and 5 October) revealed fractures of the 5^th, 6^th and 7^th ribs on the right side in a position closer to (or at) the midline rather than the spine and to the 6^th, 7^th and 8^th ribs on the left side closer to the spine than the midline. Four radiologists gave evidence but were not agreed about the timing of the fractures. Following post mortem examination, Professor Risdon noted fractures surrounded by callus on each of these ribs but also saw what he considered to be re-fractures shortly before death. As to this aspect of the case, he deferred to Professor Anthony Freemont, who is a professor of Osteoarticular Pathology at the University of Manchester.

Professor Freemont came to be called by the Crown through an unusual route. He was originally instructed on behalf of the Applicant, although for reasons that will be apparent, after he had expressed his opinion, no further reliance was placed upon him for the defence. Before the trial, however, there was a hearing in family proceedings in which he provided a report and gave evidence and Holman J subsequently made an order permitting disclosure to those concerned in the criminal proceedings of the medical evidence which had been obtained for the purposes of those family proceedings. So it was that the Crown came to learn of the evidence which Professor Freemont could give and the prosecution sought to rely on it. A defence application to exclude the evidence pursuant to s. 78 of the Police and Criminal Evidence Act 1984 was rejected.

Professor Freemont explained (and there is no dispute) that his microscopic examination of bone samples allowed him to see things not evident to radiologists who examined the distribution of calcium on an X ray. He said that there were a total of 10 different fractures dated at three different times, that is to say 4-5 weeks before death (left 6^th, 7^th and 8^th ribs), 14-28 days before death (right 6^th and 7^th ribs) and 2-5 days before death ie around the time of Rees’ admission to hospital (re-fractures of left 7^th and 8^th ribs and right 6^th and 7^th ribs with what he described as a fresh fracture of the right 5^th rib). All the experts who were asked said that the bones of a baby such as Rees are pliable and strong so that it would be necessary to exert considerable force to break them.

The force required to re-fracture previously fractured ribs (and to fracture the right 5^th rib) was the subject of detailed analysis. In his initial report (prepared for the Applicant’s solicitors), Professor Freemont said:

In his second report, Professor Freemont addressed the question whether there was any type of structural weakness in Rees’s ribs which would predispose to fracture. He found there was not. In a third report, besides summarising his conclusions, he added:

In the course of his evidence at the trial, Professor Freemont confirmed that there was nothing about Rees’s bones which indicated he was particularly vulnerable to fracture. He stated that:

In cross-examination Mr Dunkels QC focused on the fracture to the 5^th right rib, and on this witness’s conclusion that it was a fresh fracture. He put to Professor Freemont the conclusions of other expert medical witnesses that there had been an old fracture in the 5^th right rib. Professor Freemont responded by discounting that interpretation, because it was reached on x-ray evidence rather than pathological examination but, in dealing with this difference, Professor Freemont went on to address the condition of this 5^th rib in which he had noted that the medullary cavity also contained haemosiderin so that, at some stage, there had been bleeding into the bone. His report goes on:

In the course of re-examination, Professor Freemont added a further passage relevant to the condition of this rib:

Professor Freemont was clear that the issue concerning the 5^th right rib was important. He had regarded it as “the key piece of evidence …. [which] was overwhelming”. He said:

It is also clear to us from Mr Dunkels’ cross-examination that the defence team were fully aware of the significance of this issue as, indeed, was the judge. He summarised this aspect of the case in this way:

We have attempted to summarise the most important (but by no means all) the medical evidence that was called by the prosecution at the trial. Although the Applicant gave evidence, no expert evidence was called on his behalf. In the light of subsequent complaints, rather more is known about the position and it is appropriate to include the detail at this stage. As might be expected, the defence obtained reports from a number of expert witnesses, four in all. In addition to Professor Freemont, two others merely confirmed in their reports the opinions expressed by their prosecution counterparts. The fourth was Dr Waney Squier, whose report, on the face of it, was helpful to the defence and it was initially intended that she should be called as a witness at the trial. She was present in court to hear the evidence of many of the Crown’s expert witnesses, and she was able to, and did, assist Mr Dunkels with material for cross-examination. Ultimately, however, she was not called to give evidence and we deal later with a criticism of the judge’s direction in that regard.

It is sufficient at this stage to observe that Royce J provided the jury with an admirable analysis of the medical evidence (which leads us to reject a criticism made to contrary effect). Further, in the course of doing so, while later reminding the jury of the point advanced by Mr Dunkels that the case was not a trial by doctor, Royce J dealt with the medical issues in this way:

Before embarking upon a consideration of the aspects of the case upon which reliance is now placed, the lapse of time is such that it is important to provide a summary of the events following the conviction and thereby to explain why it is now, 2½ years after the conviction, that the application is being considered.

At the trial, the Applicant was represented by very experienced leading and junior counsel. After the conviction, they wrote a comprehensive advice on appeal which identified no procedural irregularity during the trial or error of law on the part of the trial judge in summing up. They did, however, express the view that the conviction for murder (rather than manslaughter) was arguably unsafe on the basis that the evidence was insufficient to prove the necessary intent: there was, so it was argued, a ‘lurking doubt’ as to that aspect of the verdict. Counsel referred to features of the evidence bearing on the issue of intent which had been emphasised in the defence closing speech to the jury: for example, evidence that the Applicant was a loving and caring parent, evidence as to the absence of any external mark of injury when Rees was admitted to hospital following his fatal collapse, expert evidence that fatal injury could be caused to a child in a sudden loss of temper with no intention to cause serious injury, and expert evidence that whoever was responsible for the fracturing of Rees’s ribs would not necessarily have been aware that such injury had been caused. Furthermore, they drew comfort from the observations of Royce J who, in his summing up, had carefully reminded the jury of those points, and had added the comment:

The single judge, Rafferty J, considered those grounds sympathetically, but on 10 June 2008 refused leave. She noted that the evidence bearing on the issue of intention to cause really serious injury, including those features of that evidence which were undoubtedly favourable to the defence, were all squarely before the jury for their consideration. Notwithstanding the force of that analysis, on 18 June 2008, the Applicant’s representatives gave notice of a renewal of the application to the full Court.

In October 2008, before the renewed application had been further considered by the full Court, the Applicant withdrew his instructions from those who had represented him at trial, and instructed a different firm of solicitors. They, in turn, instructed different leading and junior counsel. So it was that, on or about 9 January 2009, those newly instructed counsel lodged undated grounds of appeal which were presumably intended to replace those previously put forward by trial counsel. These new grounds, which extended to some 13 pages, were to the following broad effect. It was said that the jury had heard “a substantial amount of uncontested and unchallenged medical evidence” in relation to Rees’s injuries; that this was an area of medical practice in which there was a wide spectrum of professional opinion, which spectrum included the possibility of an innocent explanation of the clinical findings; and that the jury “should have had the opportunity to hear the alternative scientific arguments and then decide whether the prosecution case had been made to the appropriate standard”. It was thus submitted that “the conviction is unsafe through error and/or mistake”. In support of these grounds, on 14 January, a further report was submitted from Dr Squier; and evidence was also being sought from further experts.

The new grounds of appeal went on to assert that the failure to call Dr Squier was likely to have given the jury the wrong impression that there were no challenges the defence could have made to the Crown’s expert evidence. It was said that Dr Squier’s report “would have formed the basis of cogent evidence”, and that in failing to adduce her evidence, trial lawyers may have made an error or mistake. Complaint was also made that Dr Squier’s points had not been properly or fully explored in cross examination of the Crown’s expert witnesses.

Thus two criticisms of trial counsel were being advanced: a failure in the course of cross examination sufficiently to challenge the Crown’s experts; and a failure to adduce before the jury the cogent evidence of Dr Squier. In the circumstances and in the light of those criticisms of trial counsel and solicitors, the Applicant was invited to and did waive privilege. Mr Dunkels QC and Mr Linford provided notes dated the 17 February 2009 responding to the points made in the grounds of appeal. They have subsequently provided further information in response to a number of specific questions.

At a later date, a further undated document setting out the grounds of appeal against conviction was lodged by the same newly instructed counsel. These grounds of appeal were in some respects similar to those lodged in about January 2009, but also placed considerable emphasis on the DVD recording of Donna’s interview by the police, which feature had not been mentioned at all in the earlier grounds. It was now asserted that the DVD (a) demonstrated a potential innocent explanation for the development of both head and rib injuries; (b) had not been seen by any of the expert witnesses instructed on either side; (c) may not have been seen by defence counsel; and (d) was not seen by the jury. Thus, it was submitted, the jury had not been given:

In later correspondence, the Applicant’s solicitors said that this was not a ground of appeal based on a criticism of trial counsel for failing to make use of the DVD, but rather a ground of appeal based on the proposition that the importance of the DVD was not manifest at the time of trial. We find this distinction impossible to understand. The DVD had been made available as unused material to those representing the Applicant at trial. Also provided as unused material was a full transcript of Donna’s interview. It is clear from reading the relevant passage of that transcript that she was demonstrating to the police officers the manner in which she was saying that Rees had been thrown to the floor by his 2 year old sister. The submission that the importance of the DVD was not appreciated seems to us to be, in reality, a criticism that trial counsel either did not view the recording, or did not give sufficient weight to it, and thus failed to make use at trial of material which (it is now submitted) was important to the Applicant’s defence.

We continue with the procedural history of this application. The reports of Dr Squier, and of a further (new) expert witness Dr Cohen, were served as proposed fresh evidence. On 10 June 2009, there was a hearing for directions before the full court and a timetable laid down for the service of other evidence (which the Applicant’s advisers did not meet). Further, although 21 June 2010 was fixed for the hearing of the renewed application, on 29 April 2010, this date was also vacated. Without descending into detail, both Dr Squier and Dr Cohen had been the subject of serious judicial criticism in an unrelated case in which they had given evidence, and the Applicant’s representatives understandably felt that they could no longer rely on them as witnesses, and so sought an opportunity to instruct other experts. It should be noted that in making that application, and in response to an enquiry from the Court, leading counsel confirmed that there were no non-medical issues to be decided: a further timetable was set.

Unfortunately, this revised timetable was also not met by the Applicant’s advisers and, on 23 September 2010, a further application was made to vacate the date fixed for this application, both on the grounds that the evidence had not all been served and because of a clash in the professional commitments of leading counsel. That application was refused, with the result that the finalised expert evidence for the Applicant was only served during the last week of September 2010. As a consequence, the Crown (in the form of Mr Boyce Q.C. and Ms Martin along with the medical experts) have been forced to work to a very demanding deadline to prepare the necessary response for this hearing. In addition, new (i.e. the third) leading counsel had to be engaged on behalf of the Applicant. Mr Henry Blaxland QC was then instructed to lead Dr Thomas (who had originally replaced Mr Linford).

This change of counsel led to a further change of direction. On the 4 October 2010, the Applicant by his counsel filed further perfected grounds of appeal and a skeleton argument; these grounds differ substantially from those which were put forward in January 2009. In particular, they do not include any ground contending that the conviction for murder rather that manslaughter was unsafe on the grounds of ‘lurking doubt’ although Mr Blaxland did adopt an argument that, depending on the assessment of the new medical evidence, such a verdict could still be appropriate. We also observe that it was far from clear what ground of appeal, if any, was still being pursued in relation to the DVD recording of Donna’s interview. The only reference to that DVD formed part of a ground of appeal based on fresh expert evidence. That fresh evidence was said to include evidence to the effect:

In the light of the lack of clarity about the grounds on which this appeal has been variously pursued and bearing in mind the late date upon which Mr Blaxland became involved in the case, we have felt it appropriate to deal (albeit in some cases, shortly) with the various criticisms which have been made whether or not addressed in oral argument. Given the primacy of the medical grounds, however, we deal first with the application to call fresh evidence.

Mr Blaxland seeks to introduce new evidence from a Consultant Histopathologist, Professor Archie Malcolm, and a retired Consultant Physiologist, Dr Paul Johnson. In part, the new evidence is in the area which was covered by Dr Squier (albeit not called before the jury); in part, it is additional to that material in the sense that it deals with propositions which were unknown at the time of the trial (whether or not it was ascertainable had efforts been made to discover it). In response, the Crown have relied upon further material from Dr Stoodley, Professor Risdon and Professor Freemont while also seeking to call Dr Mark Peters, the consultant Paediatric and Neo-natal Intensivist at Great Ormond Street Hospital. We admitted all this evidence de bene esse.

In this area, Mr Blaxland sought to rely on Dr Johnson to undermine the evidence called by the Crown; although Mr Boyce Q.C. for the Crown challenged his expertise, we concluded that he did have relevant experience. His background is as a medical practitioner with early experience (in the 1960s) of neo-natology, including neonatal intensive practice. Much of this experience was in the United States, but in 1968 he returned to England as a research officer in peri-natal physiology based at Oxford University. Between 1981–2005, he held a consultancy at the Department of Obstetrics and Gynaecology in Oxford University, where he set up and ran the “Maternal Infant Healthcare and Telemonitoring Research Centre” at the John Radcliffe Hospital, Oxford.

Dr Johnson’s special interest has been the long term monitoring of infants, and research into the effects of environmental factors on foetal and post natal cardio respiratory control, into breathing disorders and heart failure. Much of his work has involved animal research and experimentation. He has no recent experience as an intensivist, nor has he ever had responsibility for extensive neo-natal or paediatric care involving acute emergencies, although he spoke of the experience of the team which he led. In evidence, he indicated that twice or three times a year, the unit for which he had responsibility would have to deal with acute clinical emergencies in children although, since the 1960s, it seems that his own hands-on experience of emergencies has been limited.

Dr Johnson explained that he could offer expertise as to the causes or potential causes of hypoxic ischaemic encephalopathy, that is to say, one limb of the “triad”. While he puts himself forward as an expert on ‘encephalopathy’ he does not even suggest that he is an expert on the causes of subdural haemorrhage. In our judgment, he was wise to disclaim such expertise in oral evidence: it was not a disclaimer he had entered at the time of any of his written reports, or indeed before giving evidence.

Before summarising the thrust of what Dr Johnson had to say on this issue, we must record our view that he was a most unsatisfactory witness. His reports (consisting of 17, 56, 4 and 13 pages) are diffuse, poorly sourced where they are sourced, hard to interpret, infused by arrogance, and quite unnecessarily combative and dismissive of other experts, including those in fields which are not his own. In order to understand what he was contending, it was necessary for us to ask him to summarise his essential thesis on two sides of A4. Discounting his definition of terms, he did so.

Dr Johnson contended that Rees had a tendency to gastro-oesophageal reflux (“GOR”) and that food stomach contents or acid entered his oesophagus. The defence mechanism to such an event is temporarily to shut down breathing, with a corresponding rise in heart rate and blood pressure (so that the brain remains oxygenated) while the contaminant is cleared. Dr Johnson suggests that in cases where the contaminant cannot be cleared quickly, or where there is an over-reaction, apnoea continues and there can be collapse accompanied by a low heart rate (bradycardia). Rees’s older sister had a reported tendency to GOR, although with no significant consequences, and we recognise that there is some historic reported anxiety by Rees’s parents about a similar problem in him, although this complaint was quite unparticularised and arose at a time after Rees’s first undeniable non-accidental injury (if that is dated by the “old” rib injuries). All that said, it is not fanciful to suggest that, at least to some extent, Rees may have suffered from GOR.

However, Dr Johnson went on to suggest that in the case of Rees, GOR was a candidate for causing his encephalopathy. On the other hand, (although it hardly needed expert evidence to tell us) it was the very clearest evidence of Drs Stoodley and Peters, and Professor Risdon, all of whom were called by the Crown to address this point among others, that GOR is extremely common. The proposition that Rees was particularly susceptible to GOR is not made out by any concrete evidence and is speculative in the extreme. Even more so, the idea that GOR led to his encephalopathy is quite unproven: furthermore, there is no proven example of GOR leading to retinal bleeding or subdural haemorrhages, even in those who have previously had subdural haemorrhage.

Further, as Dr Stoodley has pointed out, there is an incidence of subdural haemorrhage in normal babies, delivered at term, of between 9% and 46%. Thus, subdural haemorrhage must be regarded as a frequent side-effect of the stresses of delivery. Since GOR is a common phenomenon, if the normal physiological response to GOR, in the form of a spike in blood pressure, could cause in the new-born a re-bleed at the site of earlier subdural bleeding, that complication would be well-recognised. It is not well-recognised. On the contrary, it is unreported in the literature.

Doctors Stoodley and Peters, and Professor Risdon, all confirmed that they knew of no instance where the triad had been observed and was thought to be the consequence of GOR. Doctors Stoodley and Peters regularly review the images of, and care for, neo-nates who have had subdural bleeds. Neither of them has ever seen a re-bleed thought to be the result of GOR.

There is in our judgment a further crucial point. As set out in the evidence of Professor Risdon, the acute subdural bleeds sustained by Rees at the time of his collapse were at specific sites around the brain. We have set out that evidence above. These were not the sites of previous bleeding. While of course blood may migrate around the surface of the brain and down the spinal column, bleeding sites do not.

In the circumstances, we reject the explanation offered by Dr Johnson as fanciful and do not admit his evidence, the quality and tone of which did not assist us. Indeed, if advanced before a jury, his views were much more likely to obfuscate rather than elucidate. Those charged with assessing the remuneration to which he is entitled from public funds should pay careful attention to the extent to which it is directed to evidence based on his established expertise rather than argument and comment.

We must deal with one other aspect of Dr Johnson’s report, for he referred to a recent meeting organised by the Royal College of Pathologists, addressing the “triad” of evidence in traumatic head injury in children. It was convened on 10 December 2009. Present were a number of experts who have been concerned with this case including Dr Waney Squier, Dr Marta Cohen, Professor Luthert, Professor Risdon and Dr Colin Smith. This discussion did not reach a complete consensus. However there was some degree of agreement which was expressed in the paper in the following terms:

Bilateral optic nerve haemorrhage of the point where the optic point enters the globe.

Histological changes in the brain indicating mechanical damage rather than ischaemic damage (though there was not agreement on how easy or reliable such histological assessment may be).

Other post-mortem evidence of cranial trauma or extra cranial trauma.

One or more elements of “the triad” being absent.

One or more elements “of the triad being present” in a limited or a typical form …..

Young at age of death (interpreted as under 3 months, with particular caution being needed in cases young enough for birth trauma or hypoxia to be a possible explanation for the post-mortem findings)

The debate summarised in the College meeting underscores the need for caution in this difficult area. As more than one of the Crown’s experts emphasised to us, the “triad” must not be used as a tick-box exercise. In addition, the implications of the presence of the ‘triad’ must be approached with specific caution when the child in question was under three months of age, because of the risk that subdural bleeding may have been derived from the birth process. Dr Stoodley underscored this cautious approach by emphasising that it was the quality and extent of the evidence that mattered. He himself acknowledged he had seen children where there was no question of non–accidental injury, but who had some encephalopathy, some bilateral retinal haemorrhage and some subdural bleeding. Whilst the presence of all the signs in the “triad” lends support to the suggestion of mechanical trauma to the head, such findings must be placed in context, analysed and carefully considered particularly in a child as young as Rees. As this court emphasised in R v. Henderson [2010] EWCA Crim 1269, there is no room for trial by expert and in many cases the evidence including the expert evidence, may be insufficient to exclude beyond reasonable doubt an unknown cause of death. We bear these cautionary remarks very much in mind.

Professor Malcolm is the Professor of Clinical Pathology at Newcastle University and, as a Consultant Histopathologist, is (in common with Professor Freemont) an expert in bone pathology. It was common ground that this expertise (rather than that based on radiographic or other imaging of fractures) provided the best evidence on rib fractures. As Professor Malcolm puts it:

In a joint discussion prior to the hearing, Professors Freemont and Malcolm reached a fairly high degree of consensus, which was reduced into writing. Thus, it was agreed that Rees’s ribs show evidence of significant trauma occurring 4 to 6 weeks before death some of which were in the posterior segment of the ribs (6^th, 7^th and 8^th ribs on the left side) and that, other than in major accidental trauma such as a fall from a significant height or a car crash, such fractures are seen only in non-accidental injury. So it was common ground that Rees sustained non-accidental injuries to these ribs on at least one occasion 4 to 6 weeks prior to his death.

The Professors also agreed that there are fractures to the right 6^th and 7^th ribs in an antero-lateral position (on the side, roughly in line with the armpit). As Professor Risdon told the jury, these injuries are in a line, and are therefore associated. Professors Freemont and Malcolm agree that the fracture callus in these two fractures differs from that found in the left-sided injuries. However, their interpretation of this finding differs. Professor Freemont suggests the difference indicates two ages of fracture: left sided fractures at 4 to 6 weeks before death, and right sided fracture 2 to 4 weeks before death. Professor Malcolm accepts that explanation may be correct. However, he prefers the explanation that these five ribs were all injured at the same period, some 4 to 6 weeks before death, and the difference in the callus formed derives from a differential amount of movement after fracture, between the right sided fractures close to the spine (which are therefore more stable) and the left sided injuries much farther forward in the body and much farther from the spine (and therefore somewhat less stable). What is, however, agreed is that these injuries demonstrate at least one and perhaps two episodes of multiple non-accidental rib-fractures, weeks before Rees died.

In their discussions before the appeal, these two experts also agreed that the five ribs so far discussed were all re-fractured in the period 2 to 5 days before the death (that is to say, coincident with 30 September 2006). They agreed such disruption of previously fractured but healing ribs could possibly be caused by CPR but differ in their views of how likely that explanation is. Professor Freemont considered it unlikely, and he would expect more force to be required than would be produced by normal CPR; Professor Malcolm suggested this could perfectly possibly be the explanation. These experts in discussion before the hearing could only reach a broad view on this point with reference to “conventional” CPR. Their positions were refined in the course of evidence before us.

The significant disagreement between the bone pathologists is in relation to the fracture to the 5^th right rib. It is agreed this was fractured 2 to 5 days before death, in other words at a time consistent with the re-fracture of the adjacent ribs on the right, and the three ribs re-fractured on the left. We note the position of this fracture, which is in the antero-lateral position and in line with those to the other right ribs. Professor Freemont considers that the bone here fractured was normal bone, or at least bone of normal strength. In his view, this means this was clearly not caused by CPR, in part because it was a fracture of normal bone, not a re-fracture of an older healing injury. Fracture of normal bone, he says, would require more force than is applied in the course of conventional CPR: with that proposition, when he gave evidence, Professor Malcolm agreed.

Professor Malcolm’s disagreement is evident from his view that this was a recent displaced fracture through abnormal and consequently weaker bone; thus he accepts it could have been caused by recent non-accidental injury, but it could also be the result of CPR, at least if the CPR was “pretty zealous” (his words). Properly understood, therefore, the crucial issue between these experts in respect of this injury is: was the fracture of the 5^th right rib a break of bone of normal strength, and thus a recent non-accidental injury, or was it a fracture or re-fracture of abnormal, perhaps previously fractured bone, which even then might be a fresh non-accidental injury but might be the consequence of pretty zealously administered CPR?

In his report Professor Malcolm gave his conclusions on this rib in the following terms:

In oral evidence, Professor Malcolm’s key points, in addition to those summarised above, are as follows. There would have to be a significant injury short of a fracture to cause the amount of medullary bone to be found in the 5^th right rib. If the original injury to this rib had been an old healing crack fracture, he would not expect there to be any subsisting periosteal reaction by the time of examination. Although woven or immature bone is found in infants at the growth plate of a bone, he would not expect to find woven bone at this site in the middle of the bones. In giving his evidence in chief, Professor Malcolm could not volunteer any injury short of an undisplaced crack fracture which would give rise to this amount of woven bone in the medullary cavity. However, when a passage was put to him from one of Professor Freemont’s reports suggesting that there were other ways in which medullary callus might be stimulated to form, he was prepared to accept the proposition.

We have set out Professor Freemont’s primary evidence above. To this court, he restated his view that the 5^th right rib had never been fractured. In addition to the reasoning given in his earlier evidence and reports, he added one or two salient points. Damage short of fracture but so as to cause the additional bone growth could be explained by “flexing” the rib, something which could easily be consistent with fractures at an adjacent site to the other two ribs. The absence of signs of inflammation or damage in the periosteum was significant since all the “old” fractures had continuing periosteal signs at post-mortem. Professor Freemont suggested that this probably meant the 5^th right rib had never been fractured, but he added that if it had been fractured, the absence of periosteal signs meant that it was almost completely healed and therefore of essentially normal strength. Even on the latter basis, in his view this fracture could not have been caused by a normal level of CPR, or by the CPR described in this case.

Professor Freemont added that the site of this fracture was outside the normal distribution of fractures (rare though they are) caused in the course of CPR. He said the place where such fractures occur is near the anterior end of the rib which is joined to the breast bone by cartilage. Because CPR is administered by depressing the breast bone (we take it whether by finger pressure or by squeezing the chest) the biggest flex is nearest the breast bone. Here the fracture of the 5^th right rib (and the fractures to 6^th and 7^th ribs) is to the side, in line with the armpit.

In answer to Professor Malcolm’s point that there was a high amount of woven (as opposed to lamellar) bone at this point in the 5^th rib, Professor Freemont pointed out in his view there was also a high level of woven bone in the cortices of the 6^th and 7^th ribs. We understand the significance of this to be that, despite that finding, all the experts were agreed that the original fractures of those ribs were definitely caused by the use of unlawful force. Regrettably, this point was never put to Professor Malcolm.

Taking Professor Malcolm’s view at its highest, therefore, in order to exclude his possibility that the fracture of the 5^th right rib might have been caused by the CPR administered either by the Applicant or the staff nurse on admission (the full descriptions of which we have set out above), consideration has to be given to the question whether that could be described as “pretty zealous”. In that regard, it emerged that Professor Malcolm had never been shown the evidence concerning the handling of Rees by the Applicant and the staff nurse attempting resuscitation in hospital, (putting him in a different position from Professor Freemont, who saw and heard the witnesses). The evidence given by the applicant was put to Professor Malcolm and he agreed that description sounded “mild”. From the description given, it sounded to Professor Malcolm that it would be unusual for that to cause a re-fracture never mind a fresh fracture. He felt that the description given “did not sound as though very significant pressure had been used”. Mr Blaxland invited us to consider that there might have been a natural tendency on the part of a father charged with murder to diminish the description he gave of the force used. We do not understand why that should be so, given that he certainly knew by the time he gave evidence that Rees had suffered recent fractures of his ribs.

As to the description by the staff nurse, he believed that the bruise demonstrated that the attempts at CPR had been “clearly reasonably forceful”, but Professor Malcolm agreed that since the pressure was on the centre of the chest, he would not expect such CPR to cause this injury. He did point out that although the CPR was mounted at a site relatively remote from the fracture site to the 5^th right rib, pressing on the centre puts stress all around the rib cage.

That was not the only evidence about the force required to cause the rib injuries that we heard. In addition to the views of Professors Freemont and Malcolm, Professor Risdon (who deferred to the histopathologists on interpreting fractures but not necessarily on causation) said that almost every child upon whom he had performed a post mortem had been extensively resuscitated. He went on:

We also heard evidence relevant to the question of CPR from Dr Mark Peters, the Consultant Paediatric and Neo-natal Intensivist at Great Ormond Street Hospital. Dr Peters has worked in tertiary paediatric intensive care since 1994 and is involved in the care of 1,500–2000 episodes of critical illness in children every year. He and his team have very frequent, literally hands-on, involvement in resuscitation of critically ill neo-nates and young babies. Dr Peters emphasised to us the necessary vigour of resuscitation. He said that he himself was pretty fit, but was unable to sustain appropriately vigorous resuscitation for more than about 5 minutes, and hence resuscitation was conducted by a rotating team of staff. In his unit, staff are taught that two finger resuscitation of the kind administered by the staff nurse in this case is not regarded as optimally vigorous and the preferred method is to resuscitate by squeezing the chest of the infant with both hands. Dr Peters has done that on very many occasions and never broken a previously unbroken rib. He has performed that kind of resuscitation on numerous patients who already have fractured ribs, perhaps mostly ribs surgically fractured for the purpose of cardiac or other operations. This evidence gives some important context to the clash of opinions between the histopathologists as to the potential impact of the CPR administered to Rees and adds colour to Professor Malcolm’s description of what must be “pretty zealous”.

We draw together some of the threads concerning the rib injury as follows: Professor Freemont and Professor Malcolm were both extremely well qualified experts who at all times did their best to help the Court. The pathologists are undoubtedly better positioned than any other expert in deciding whether the 5^th right rib had, or had not previously been fractured. There is a genuine professional difference between the two on the point, although each accepts that the other might be right. There was clearly some previous abnormality in the 5^th right rib but unless the rib was weakened, all are agreed it would not be broken by any CPR, even zealous CPR.

In the end, Professor Malcolm is not able to say with confidence that the presence of woven bone at the relevant site can only be explained by inferring a previous crack fracture. The alignment of the fracture site in the 5^th right rib with the acknowledged previous fractures in the 6^th and 7^th ribs would tend to associate the previous episode of damage to the 5^th right rib with the same incident which caused those earlier fractures. However, the alignment does not serve to distinguish a previous undisplaced crack fracture from a previous flexing or stressing of the rib causing damage short of fracture.

There is one further element of the picture to which it is appropriate to refer before considering the effect of this evidence. Professor Risdon made the valid point:

The material governing principles are set out in s. 23 Criminal Appeal Act 1968, as amended by the Criminal Appeal Act 1995, (“the 1968 Act”) in these terms:

…

…

receive any evidence which was not adduced in the proceedings from which the appeal lies.

The Court of Appeal shall, in considering whether to receive any evidence, have regard in particular to─

whether the evidence appears to the Court to be capable of belief;

whether it appears to the Court that the evidence may afford any ground for allowing the appeal;

whether the evidence would have been admissible in the proceedings from which the appeal lies on an issue which is subject of the appeal; and

whether there is a reasonable explanation for the failure to adduce the evidence in those proceedings.”

The ambit of the provision is most clearly expressed in R v Steven Jones [1997] 1 Cr. App. R. 86 by Lord Bingham CJ when he said (at 92G):

This latter difficulty was also discussed in R v. Kai-Whitewind [2005] 2 Cr App R 31, R v. Meechan [2009] EWCA Crim 1701 and R v. Henderson [2010] EWCA Crim 1269 the latter of which underlined that a conviction can only be regarded as safe if the case proceeds on a logically justifiable basis, including the acceptance or rejection of expert evidence, that evidence may be insufficient to exclude, beyond reasonable doubt, an unknown cause and that, although appeals should present an opportunity to call new experts in the hope that they might do better than those who had been rejected or (as in this case) whom the defence had chosen not to call, it is difficult to exclude new medical evidence which is both cogent and relevant on the basis that it should have been called at trial (see paras. 1 and 3 per Moses LJ).

Thus, considering those principles, the first stage is to consider whether evidence heard de bene esse should, in fact, be received either because it satisfies the criteria set out in s. 23(2) of the 1968 Act (which, in the context of this case concerning expert evidence is likely to mean whether there is a reasonable explanation for the failure to adduce the evidence at trial), or whether it is appropriate to go beyond the particular considerations set out in s. 23(2) to the wider interests of justice as left open by s. 23(1) and clearly relied upon in Henderson. If a decision is made to receive the evidence, the second stage is to determine whether it appears to the Court that the evidence affords any ground for allowing the appeal on the basis that although contained within s. 23(2), if it does not afford a ground for allowing the appeal, the material takes the case no further.

That brings the court to define the grounds for allowing an appeal on this basis, the principles of which are set out in a number of authorities at the forefront of which is R v. Pendleton[2001] UKHL 66; [2002] 1 Cr. App. R. 34; [2002] 1 WLR 72 (per Lord Bingham of Cornhill, at page 83, paras. 18 and 19) which was followed by this court in R v. Hakala [2002] EWCA Crim 730 and R v. Hanratty [2002] EWCA Crim 1141, [2002] 2 Cr App R 30. This line of cases was cited in Dial & anorv. State of Trinidad and Tobago [2005] UKBC 4; [2005] 1 WLR 1660 by Lord Brown of Eaton-under-Heywood who gave the judgment of the majority (the others being Lord Bingham of Cornhill and Lord Carswell) and put the matter in this way:

That approach was adopted in R v. Harris [2006] 1 Cr App R 5 (see per Gage LJ at para. 101) and it has been recognised that the “emphasis [in Dial]was put in a slightly different way” (R v. Dunn & ors [2009] EWCA Crim 1371 per Goldring LJ at para. 111). Mr Blaxland, however, went further and submitted that Dial weakened the rigour of the test identified in Pendleton as exemplified by the fact that Lord Steyn (who had been party to the decision in Pendleton) dissented. In fact, Lord Steyn did not specifically challenge Lord Brown’s analysis of the law but strongly disagreed with the majority on the approach to the specific facts of the case. Thus, the majority took the view that the Court of Appeal of Trinidad and Tobago had been entitled to conclude that a proved significant lie by the main identifying witness was not so central to his identification as to throw real doubt upon it and so render the verdicts unsafe. Lord Steyn (and Lord Hutton) concluded that the Court of Appeal had not considered how the jury would have viewed the potential impact of the lie on the overall truthfulness of his evidence. They also identified further reasons which they considered undermined the safety of the convictions.

Furthermore, as to the principle, it is important to underline that Lord Bingham was part of the majority judgment articulated by Lord Brown and would hardly have been so had he considered that the analysis of Pendleton and the subsequent decisions was not both accurately reflected and fairly illuminated by Lord Brown’s exposition of the law. We have no doubt that it was: both in Stafford v DPP and Pendleton, the House of Lords rejected the proposition that the jury impact test was determinative, explaining that it was only a mechanism in a difficult case for the Court of Appeal to “test its view” as to the safety of a conviction.

As Lord Bingham made clear in R v Steven Jones, the scheme of s. 23 of the 1968 Act is not designed with expert evidence in mind not least because Professor Malcolm was clearly available at the time of the trial but, having received an adverse opinion from Professor Freemont, it is understandable why it was felt that it would be inappropriate to instruct a further expert. In that regard, ‘expert shopping’ is always to be discouraged and although public funds should always be available to instruct suitable experts when defending a criminal prosecution, it is less clear whether such funds should continue to be expended in seeking further expert evidence (in the hope of finding someone to say something different) after conviction. Thus, in this case, four experts were instructed at the time of the trial; at least six experts (five of whom were new to the case) have provided further evidence for the Applicant’s present advisers, of whom two have been called. On the other hand, if fresh credible expert evidence has, in fact, been obtained which does provide a real argument as to the safety of a conviction, it is almost inconceivable that the court should not fully consider that evidence and its implications, not least to avoid the risk of what would be a miscarriage of justice. Ultimately, the court will evaluate the material placed before it: this means that it must be for the authority responsible for the provision of public funding to set out clear guidelines as to the circumstances in which new experts may be instructed after conviction and for steps to be taken to ensure that, whether an approach is to the Legal Services Commission or the Registrar of Criminal Appeals, common standards are understood and applied.

As for this case, there is no doubt that Professor Malcolm provided admissible evidence which is capable of belief; the real issue, therefore, is to evaluate its importance in the context of the remainder of the evidence in the case and thus determine whether it renders the verdict of the jury unsafe. Considering first the fifth rib, we recognise Professor Malcolm’s theoretical analysis of the strength of the 5^th right rib on 30 September. Putting Professor Freemont’s view to one side, it remains necessary to consider the force used in CPR (which was a question left by Royce J to the jury). Professor Malcolm did not suggest that the Applicant’s description of the CPR performed on Rees could possibly account for injury and, although he took a slightly different view in relation to the staff nurse (while accepting that this fracture required more trauma to cause it than the others without suggesting that it did), the evidence of Professor Risdon and Dr Peters, speaking from their different perspectives, in our judgment, moves Professor Malcolm’s possibility into the realm of the theoretical rather than the realistic.

That is not all, because it is also important to consider the contextual facts. These are as follows. First, Rees was born by elective Caesarean section such that it was extremely unlikely he suffered birth injury and, in any event, he exhibited no signs thereafter or during his short life of having sustained birth injury. Second, Rees saw numerous (and different) medical professionals over the weeks. Although there were some concerns, save for the GOR (which we have rejected) it has not been suggested that any of his other symptoms suggested an underlying pathology (and, in particular, a neurological pathology). Third, at what at the latest was 4 weeks of age, it is beyond doubt that Rees suffered multiple rib fractures as a consequence of non-accidental injury: the incident with his sister simply could not have caused the injuries that he had sustained. There is also historic subdural bleeding, which, although theoretically consistent with the time of birth (rendered even more theoretical because birth was by caesarean section), is much more easily associated with the known earlier non-accidental injury. Fourth, Rees behaved entirely normally throughout 30 September: he fed normally and appeared content, capable of following the movement of Mr Randal’s hand. Fifth, if the Applicant’s account is to be believed, Rees continued to be normal and was so when he sent the text to his wife at 7.44 pm asking where she was; it was only thereafter that he noticed that Rees was clearly unwell and by 8.12 pm he was at the hospital; if that is right, the deterioration was extremely sudden. Sixth, even on Professor Malcolm’s most favourable thesis, there is re-fracturing of six ribs which are easily compatible with further non-accidental injury, it only being, as we find, theoretically possible that it was associated with CPR as described at the hands of the Applicant and the staff nurse.

Returning to the test, although we accept that Professor Malcolm provides a new consideration of the injury to the 5^th rib, we reject the submission made by Mr Blaxland that this, of itself, is sufficient to quash the conviction on the grounds that had he been called to give evidence, the jury may have had some doubt about the conclusions of Professor Freemont that CPR could be excluded as a cause of all the re-fractures. In our judgment, on the issue of the use of unlawful force on Rees (a requirement for any conviction), we do not accept that this new evidence impacts on the circumstances sufficiently to affect the decision of the jury on the issue of whether unlawful force was used. Paying full attention to the strictures in Henderson to which we have referred, in the light of all the circumstances, we consider the possibility that this triad of injuries arose from encephalopathy of unknown cause (whether associated with GOR or not) to be so vanishingly unlikely that it can be rejected entirely; furthermore, this conclusion is not affected by the theoretical possibility, as we find it to be, that the 5^th right rib might not have required as much force to fracture it as Professor Freemont believes. To modify Professor Risdon’s observation as to the approach of doctors, this is the effect of taking the components of the incidents and analysing the clarity of the picture that emerges.

That is not a complete end of the matter because Mr Blaxland also argued that even if his primary submission was to be rejected, given the uncertainty about the degree of force necessary to cause the triad and the fact that the evidence of the degree of force used on 30 September was heavily dependent on Professor Freemont’s conclusions as to the 5^th right rib, even if the evidence of Professor Malcolm did not impact on the validity of the conclusion as to unlawful force, it did impact on the issue of intent with the result the conviction for murder is unsafe and a conviction for manslaughter ought to be substituted. Mr Boyce Q.C. argued that given the unlawful use of force on 30 September leading to injury to the ribs, it was an inevitable inference that the Applicant had also caused the earlier fractures which certainly required considerable force with the result that it was open to the jury to convict of murder.

It is common ground that it is not possible to reach a conclusion as to the degree of force likely to be necessary to cause the triad of injuries which ultimately led to Rees’s death. The prosecution case (as summarised by the Judge) was that the unlawful act “was likely to have been shaking or a shake plus throwing Rees against a soft surface or simply throwing against a soft surface” Save for Dr Markham (who spoke of trauma such as throttling or a blow) the balance of opinion (expressed by Dr Stoodley) was that the injury resulted from a shaking mechanism of a type that “the degree of force applied to a two month old child would be obviously inappropriate”.

It was this that led Royce J to deal with intent in the way that caused counsel initially instructed to mount an appeal. As the conclusion of his summing up, he dealt with Mr Dunkels' speech in this way:

It is not suggested that Mr Dunkels’ summary of the evidence was not accurate which generates the question: what was the evidence from which the jury could infer intent to cause really serious injury? In reality, it was the possible inference that could be drawn from the ‘shaking or throwing’, together with Professor Freemont’s evidence that considerable force was required to fracture the 5^th right rib which he considered to have been effectively contemporaneous with whatever injury caused the triad. We have rejected the proposition that this fracture could have been caused by CPR and concluded that there is no basis for interfering with the finding of unlawful force but that is not to say that Professor Malcolm’s evidence might not have affected the conclusion to be drawn as to manner in which the unlawful force was inflicted. Given the paucity of the evidence as to the necessary force to cause the triad, there is a risk that the jury would have used the emphatic degree of force of which Professor Freemont spoke in relation to the right 5^th rib as a proxy to determine the Applicant’s intention at the time of shaking or injuring Rees in such a way as to cause the triad. As Professor Malcolm had ‘softened’ the extent to which emphasis could be put on the fact that the right 5^th rib was of normal strength, the question is whether the inference of intent may have been approached differently. Although we have some real hesitation, we are not prepared to conclude that the conviction for murder (as opposed to a conviction for manslaughter) remains safe. In that regard, we are particularly mindful that the original defence team had mounted an appeal on the basis of ‘lurking doubt’ and although that ground, on its own, would not have caused us to take a different view, the gloss provided by Professor Malcolm plays into the same issue.

Before reaching a final conclusion, we must turn to the other points raised both in the perfected grounds of appeal, and in earlier grounds, not all of which were pursued in oral argument before us, although none appeared formally to have been abandoned.

The first concerns the criticism of the approach in cross examination to the Crown’s expert witnesses. Both from an analysis of the judge’s summing up, a consideration of those transcripts which we have seen and a view of some of the material available to Mr Dunkels, we have no doubt that Mr Dunkels did cross examine prosecution experts in those areas in which it was legitimate to believe that progress could be made. Dr Squier was, of course, present in court during the evidence of a number of the Crown’s experts, was plainly able to assist counsel in his cross examination and clearly did so. Without needing to go into detail, and without confining ourselves to the matters which were within Dr Squier’s field, we note that the matters raised in cross examination included the possibility that ribs may have been fractured or re-fractured during the attempts at resuscitation on 30 September; possible explanations other than trauma for intracranial bleeding, such as a subdural haemorrhage existing at birth, or reperfusion through damaged capillaries when Rees’ breathing was restarted at hospital; and the existence of controversy within the medical profession as to the significance which should be attached to a finding of the triad. It seems to us that Mr Dunkels put all relevant points, and in doing so achieved as much success as could realistically be expected given the strength of the medical evidence in support of the Crown’s case.

Those now appearing for the Applicant rightly do not pursue the criticisms as to the adequacy of cross examination. We recognise that counsel who settled the new grounds of appeal had not had an opportunity to see the transcripts which have subsequently been obtained. In those circumstances, we merely say that the criticisms which were initially made are without substance, and afford no possible basis for regarding the conviction as unsafe.

The second ground which we consider is based on the fact that Dr Squier was not called to give evidence is concerned. In this regard we are assisted by the notes which have been provided by Mr Dunkels QC and Mr Linford. These notes incorporate the relevant passages from contemporaneous attendance notes made by their instructing solicitors. It is entirely clear that that there was good reason for the decision which was taken not to call Dr Squier to give evidence.

It had been the intention of those representing the Applicant at trial that Dr Squier would be called as an expert witness. Indeed, at one point in his cross examination of Professor Risdon, Mr Dunkels said, with reference to Dr Squier, “I think when we hear from her she will tell us…” The notes indicate however that the intention of calling her was abandoned as a result of a consultation with Dr Squier after the Crown’s experts had completed their evidence. Dr Squier then indicated that she would not now be in a position to give evidence which would assist the Applicant’s case. It was apparent moreover that if cross examined about certain points the answers she would have to give could be positively damaging to the Applicant’s defence. Crucially, Dr Squier indicated that she was unable to reconcile the combination of the brain injuries, the rib injuries, and the Applicant’s account of the events of 30 September 2006. Had she been cross-examined about that combination – as was highly probable, since it lay at the heart of the Crown’s case – her evidence was likely to have been particularly damaging to the Applicant’s case. The Applicant was advised to that effect, and it was agreed that Dr Squier should not be called.

In short, Dr Squier was not called because it became clear during the trial that she could not assist, and might well damage, the Applicant’s case. On the basis of counsel’s notes and the solicitors’ attendance notes, it cannot possibly be argued that the decision not to call Dr Squier was in any sense an error. On the contrary: it seems to us that the advice given to the Applicant by trial counsel in this regard was plainly correct.

The contention that trial counsel was wrong not to call Dr Squier as a witness has also not been pursued before us, though we understand that Dr Squier is said to have expressed a different recollection of what she said during the consultation. In our view, and notwithstanding any difference of recollection on Dr Squier’s part, it is regrettable that this criticism was ever advanced at all. Guidance issued by the Bar Council sets out the formal procedures to be followed where grounds of appeal involve criticisms are made of former counsel. It concludes by saying (see C-45 Archbold, Supplement at para. 7):

The third ground of appeal which we consider concerns the DVD recording of Donna’s interview. We can deal with this shortly. In the absence of Donna from the witness box, the demonstration which she gave to the police officers of the manner in which Rees was said to have been thrown to the floor was hearsay. It was not accepted by the Crown as truthful or accurate. If the defence had wished to put that demonstration forward as an accurate representation of what had happened, so as to provide a basis for an expert witness to opine on issues in the case, a hearsay application would have been necessary. Given that Donna could have been called to give the evidence herself, we see no basis on which it could be argued that such a hearsay application could have succeeded. No such basis is mentioned in the written grounds of appeal.

As an alternative, it is submitted in writing that the DVD recording is admissible as fresh evidence. Given that it had been made available to the Applicant’s representatives in advance of the trial as unused material and that the fact that the interview had been recorded on DVD and the fact of the demonstration, were both apparent on the face of the transcript, it is simply not open to argument that it was not available for use at the trial.

Finally and for the avoidance of all doubt, it is now clear from the expert evidence that we have heard that the multiple fractures of Rees’ ribs could not be explained by reference to the incident that Donna described: the DVD does not advance the case of the Applicant at all.

The next ground of appeal is based on the assertion that background evidence of a highly prejudicial nature was wrongly admitted at trial. The evidence in question was to the following effect. Firstly, that in late 2004 or early 2005 the Applicant’s stepmother had suggested he should attend an anger management course. The Applicant accepted that he would shout and swear at other motorists when he felt he had been “cut up” whilst driving. He went to an assessment session and then attended one session with a counsellor. Secondly, that for a short period in 2005 the Applicant had a brief sexual affair with a woman who was Donna’s cousin, and that there had been many arguments between the Applicant and Donna when she learned about this affair. The ground of appeal contends that this evidence could only ever have been admissible as evidence of bad character under s101(1)(d) of the Criminal Justice Act 2003. No bad character application was ever made, and (it is submitted) any such application would have been bound to fail because the evidence could not possibly be regarded as showing a propensity for violence towards a child, and so could not satisfy the criteria for admissibility in s101(1)(d), as supplemented by s103 of the 2003 Act.

We see no merit in this ground of appeal. An important part of the Applicant’s case at trial was to emphasise his positive good character, relying not only on the fact that he had never been convicted of any criminal offence but also on his exemplary record of military service. He also wished, understandably, to rely on the favourable observations of a number of witnesses as to his treatment of his two children, the general view of those witnesses being that the Applicant was a loving and caring father. Moreover, it was clear that if the jury were satisfied that someone had caused unlawful injury to Rees, the only realistic candidates were the Applicant and his wife. Another important part of the Applicant’s case, accordingly, was to draw attention to matters which might be regarded by the jury as implicating Donna rather than himself. Examples of such matters were the inconsistencies in the accounts which Donna had given of the incident between Rees and his sister, and evidence that she had at relevant times been suffering from depression, from initial difficulties in bonding with Rees after his birth, and difficulty in coping with him at home. We therefore do not agree that s101(1)(d) of the 2003 Act was the only possible gateway to admissibility if the Crown were required to make an application to adduce bad character evidence: s101(1)(g) was also relevant because an attack was being made on Donna’s character.

In those circumstances, it came as no surprise to this court to receive from Mr Dunkels written confirmation of that which we had in any event assumed to be the position: namely, that the evidence adverse to the Applicant was admitted by agreement, pursuant to s101(1)(a) of the 2003 Act. Mr Dunkels took the view that it would be advantageous to the Applicant to have the evidence of his good character before the jury, even though that would probably result in the Crown being entitled to adduce the evidence as to the anger management course. We agree with that analysis: the decision of Mr Dunkels to reach agreement with prosecuting counsel, rather than requiring the Crown to make a bad character application which could be expected to succeed, was in our view an entirely sensible one, and certainly not one which can legitimately be criticised in this application.

So far as the evidence of the extra-marital affair is concerned, Mr Dunkels has indicated that he and his junior took the view that it would assist the Applicant’s case for this evidence to be before the jury. Such evidence could help to explain the family dynamics, and to provide an explanation for why some members of the family might be ill disposed towards him in their evidence. Given the circumstances which we have summarised above, and in particular the fact that Donna provided the only other plausible candidate for the inflicting of any non-accidental injury, that seems to us to be an entirely understandable view for counsel to have taken. Again, it is certainly not an aspect of their conduct of the case which could possibly give rise to an arguable ground of appeal. To be fair to Mr Blaxland, once he had seen Mr Dunkels’ explanation, he did not seek to pursue these grounds orally.

We add for completeness that, in any event we see no merit in the suggestion that the jury might have been unfairly prejudiced against the Applicant by this evidence. The anger management course was something which he attended on one occasion only, principally because members of his family were concerned about his verbally aggressive attitude towards other motorists when driving. The extra marital affair was of short duration, plainly the subject of regret and profuse apologies to his wife, and could not sensibly be regarded as undermining his denial that he had inflicted any unlawful injury on his son. In short, we take the view that these were matters which had to be in evidence if the Applicant was going to present the positive side of his good character, and to put forward the possibility that Donna had inflicted the injuries on Rees, but which cannot in any event have caused any weakening of his defence. There is nothing in this ground of appeal which casts any doubt on the safety of the conviction.

We turn finally to a ground of appeal relating to the summing up. We have already considered the criticism that the judge failed to assist the jury as to aspects of the medical evidence. There remains for consideration a complaint that the judge “commented in trenchant terms about the fact that the defence had failed to call any expert evidence”. It is submitted that that comment “is likely to have led the jury to conclude that the defence had no answer to the prosecution’s experts and this evidence proved guilt”.

In his closing speech for the Crown, Mr Ford QC (who then appeared for the prosecution) had commented to the jury about the fact that Dr Squier had been present at court but had not been called as a witness. Mr Dunkels, in his closing speech for the defence, dwelt upon the differences of opinion between the Crown’s medical experts, and with reference to Dr Squier he asked rhetorically whether it would really have assisted the jury to hear a yet further opinion. Thus both prosecution and defence had made specific reference to Dr Squier’s absence from the witness box.

At an early stage of his summing up the judge directed the jury about expert witnesses in conventional terms. In that part of his summing up he said –

Before closing speeches there had been the customary discussion between the judge and counsel as to matters of law. The judge had indicated that he would give a direction to the effect which we have just quoted. No objection was raised by counsel at that time, and that passage in the summing up is not the subject of any criticism before us.

Complaint is however made of a passage later in the summing up, at a stage when the judge was reminding the jury of some of the closing submissions of counsel. Referring to Mr Dunkels, the judge observed –

Mr Blaxland in his submissions to us has pointed out that the judge made very few comments in a summing up which was commendably succinct. This comment, he submits, was unnecessary, and “may have made a real difference”.

We see no substance in the criticism made of that judicial comment. Firstly, it seems to us that it was perfectly proper for the judge to mention the point, since both counsel had addressed the jury about it. Secondly, his comment was a legitimate one in the light of what had happened in the course of the trial: Dr Squier had plainly been identified in the presence of the jury as an expert witness who was not merely there to assist with cross-examination but would also be giving evidence, and yet when the appropriate time came she was not called. The jury were therefore likely to wonder why that was so, and the judge was in our view correct to deal with the point. Thirdly, the judge made it perfectly clear that these were all matters for the jury, and immediately went on to reiterate his direction that the burden of proof was on the prosecution. He followed that direction with a reminder of Mr Dunkels’ overall submission as to the insufficiency of the prosecution evidence. For those reasons we reject the submission that the jury’s verdict is rendered unsafe by reason of this comment.

In the circumstances, we grant the Applicant leave to appeal against conviction only on the basis of the new evidence adduced from Professor Malcolm and we reject all the remaining grounds. Given our view that the conviction for murder is unsafe but that it would be appropriate to substitute a conviction for manslaughter, we invited the Crown to consider whether it sought to apply for a retrial. In the event, Mr Boyce made it clear that, prior to the trial, although satisfied that the relevant test for prosecuting the crime of murder had been met, the Crown had been willing to accept a plea of guilty to manslaughter. In the light of our conclusions, however, together with the advantage of hearing Professor Malcolm and testing his evidence, the view was now taken that the evidential test in the ‘Code for Crown Prosecutors’ is no longer satisfied. Thus, Mr Boyce did not seek a retrial: we agree with that conclusion and substitute for the conviction for murder a conviction for manslaughter.

The sentence of life imprisonment with a minimum term of 13 years falls away and the Appellant (as he has now become) now falls to be sentenced for manslaughter. We say immediately that the range of sentences imposed for manslaughter of a small child is wide. This is because there are very real variations in the gravity of the conduct that leads to death and, furthermore, the intent leading to a conviction for manslaughter also covers a broad range, from the intent to commit a minor assault up to an intent falling just short of that necessary for murder. The assessment of these features is obviously highly relevant when considering culpability in a given case. As must be clear from our conclusions when quashing the murder conviction, our view here is that the evidence points, at least, to an intent just short of that necessary for murder

In this case, the aggravating features are the extreme vulnerability of the victim, Rees. Further, although Mr Blaxland submitted to the contrary, in our judgment, Royce J was entitled to conclude (as we conclude) that the Appellant had caused the earlier fractures to the ribs on one if not two occasions and that to do so required considerable force. Neither is it irrelevant that these earlier fractures had been re-fractured; doubtless, this did not require the same force but we reject the view that we must proceed to sentence on the basis that these injuries were all caused by CPR. For the reasons that we have given, whereas there might be an issue as to the extent of the force applied, the overall circumstances are such that we have no doubt that they were not accidental in origin.

That is not the extent of the elements aggravating this killing. Although not mentioned by Royce J, it is legitimate to infer that the fatal injury was inflicted in temper, about which the Appellant had been warned by his family and advised to seek help. He was not honest in what he told the hospital and, of greater significance, maintained his denial in circumstances where his wife was necessarily implicated in the earlier injuries. On the other hand, we also reflect upon the mitigation which the judge accepted: there was no intent to kill, no premeditation and the incident, in temper, was committed against a background of stress in his marriage with him working long and unsociable hours. The judge also took into account (as do we) the Appellant’s good character and exemplary record in the army.

Turning to the guidance which can be obtained from the decisions of this court, Mr Blaxland submitted that the bracket was 5-7 years. He referred to R v Yates [2001] Cr App R (S) 428 in which a sentence of 7 years was reduced to 5 years following a plea to the offence of manslaughter following shaking of an 11 week old baby who had also suffered a fractured skull. He also referred to the unusual case of R v Owen [2009] EWCA Crim 702 in which the sentence of 3 years was passed following an admission of causing grievous bodily harm contrary to s. 20 of the Offences against the Person Act 1861 to a 6½ month old child involving a triad combination and fractures to an arm which caused cerebral palsy but not immediate death. When the child died 7 years later, the plea of guilty to manslaughter led to an additional sentence of 12 months, the trial judge noting that the appellant had rebuilt his life since his earlier release from prison. In our judgment, the facts of this case are so unusual that no assistance can be derived from it: in any event, Hallett LJ made the point that some may argue that a sentence of 4 years imprisonment imposed had the child died immediately “would not be long enough”.

It is clear that these cases fell to be decided without regard to the Criminal Justice Act 2003 (“the 2003 Act”); even if, technically, Owen concerned a death subsequent to the commencement of the Act albeit that the unlawful act was many years prior to it. What the 2003 Act did was to increase the punitive element in the sentence for murder. In R v Wood [2010] 1 Cr App R(S) 2, Lord Judge CJ (giving the judgment of a five judge constitution dealing with a case of manslaughter by reason of diminished responsibility) acknowledged that there was no express link between the guidance in Schedule 21 and the principles applicable to sentencing decisions in such cases but pointed out that the reduction of a defendant’s mental responsibility for a killing did not alter other circumstances of the homicide. He then said:

Similarly, in another five judge constitution, Attorney-General’s Ref No. 60 of 2009 (Appleby and others) [2010] 2 Cr.App.R.(S) 46 (“Appleby), Lord Judge CJ considered the appropriate sentence for unlawful act manslaughter albeit in the context of incidents of public disorder. He recognised that the maximum penalty for inflicting grievous bodily harm contrary to section 20 of the Offences against the Person Act 1861 was 5 years imprisonment but set the scene by observing (at para 3):

Having considered the authorities and referred back to Wood and the disparity between a sentence for murder with the tariff fixed by reference to Schedule 21 of the 2003 Act and that for manslaughter, he went on (at para 22)

When asked to address the impact of Schedule 21 to the Criminal Justice Act (“the 2003 Act”) on homicide short of murder, Mr Blaxland relied on s. 143(1) of 2003 Act (to the effect that the court must consider the offender’s culpability in committing the offence and any harm which the offence caused) and para. 1.7 of the guideline issued by the Sentencing Guidelines Council on Overarching Principles – Seriousness to the effect that harm must always be judged in the light of culpability. He referred to the fact that the guideline on manslaughter by reason of provocation contained no reference to the need to increase substantive offences for manslaughter in order to reflect the increase in the minimum term for murder and submitted that it would be inappropriate to allow Schedule 21 to impact on sentences which flow from the death of a child in circumstances such as here exist. As Lord Judge explained in Appleby, however, this guideline was the culmination of work of the Sentencing Advisory Panel and Sentencing Guidelines Council which began before the 2003 Act was brought into force and, although it was referred to, this specific feature was not considered.

The 2003 Act prescribes the starting point of 15 years for murder which is neither exceptionally nor particularly high and identifies as an aggravating feature the fact that the victim was particularly vulnerable because of age (see paras. 6 and 10(a) of Schedule 21). It is doubtless for that reason that, notwithstanding the mitigation that he accepted, Royce J imposed the inevitable mandatory life sentence with a minimum term of 13 years and it has never been suggested that this sentence was either wrong in principle or manifestly excessive. The offence of manslaughter is, of course, of a very different order of gravity (there being no intention to cause really serious harm) but the vulnerability of the victim is also clearly imported into the offence of manslaughter (see the Definitive Guidelines on Overarching Principles in relation to Assaults on Children at para. 9 and in relation to Seriousness at para 1.23). In the light of this background and all the circumstances of the case, we have no doubt that the bracket suggested by Mr Blaxland is simply too low.

In our judgment, for the reasons expressed in Wood and Appleby, the court must reflect, at least in part, the increased focus on the fact that a victim has died in consequence of an unlawful act of violence. Although, as Mr Blaxland argues, we recognise that this Appellant has had to suffer not only the loss of his own child, but must also come to terms with the guilt of having caused the child’s death, the very least sentence that we believe it appropriate to impose is one of 10 years imprisonment. The period of 7 days spent in custody prior to sentence being imposed by Royce J will similarly be taken into account.